diff options
author | Michael Niedermayer <michaelni@gmx.at> | 2012-08-03 22:43:44 +0200 |
---|---|---|
committer | Michael Niedermayer <michaelni@gmx.at> | 2012-08-03 23:13:06 +0200 |
commit | a7acab6cda69d55c3d9d75a3ebb6cff68b15e689 (patch) | |
tree | fb79107f7f9e93451092d1f246243b9022a9ee2e /libavcodec/g723_1.c | |
parent | a763cafc0c69e3fad91f97867b942182804f79b0 (diff) | |
parent | 9cc74c9f6e8b645e67d45b2070db004caca09af7 (diff) | |
download | ffmpeg-a7acab6cda69d55c3d9d75a3ebb6cff68b15e689.tar.gz |
Merge remote-tracking branch 'qatar/master'
* qatar/master:
vc1dec: Remove separate scaling function for interlaced field MVs
vc1dec: Invoke edge_emulation regardless of MV precision
x86: Use consistent 3dnowext function and macro name suffixes
g723_1: scale output as supposed for the case with postfilter disabled
g723_1: increase excitation storage by 4
g723_1: fix upper bound parameter from inverse maximum autocorrelation
g723_1: make scale_vector() behave like the reference
g723_1: fix off-by-one error in normalize_bits()
g723_1: save/restore excitation with offset to store LPC history
wmapro: prevent division by zero when sample rate is unspecified
x86: proresdsp: improve SIGNEXTEND macro comments
x86: h264dsp: K&R formatting cosmetics
LICENSE: Document all GPL files
Conflicts:
libavcodec/g723_1.c
libavcodec/wmaprodec.c
libavcodec/x86/h264dsp_mmx.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'libavcodec/g723_1.c')
-rw-r--r-- | libavcodec/g723_1.c | 28 |
1 files changed, 19 insertions, 9 deletions
diff --git a/libavcodec/g723_1.c b/libavcodec/g723_1.c index c7f6ac9301..80d1a04698 100644 --- a/libavcodec/g723_1.c +++ b/libavcodec/g723_1.c @@ -52,7 +52,7 @@ typedef struct g723_1_context { int16_t prev_lsp[LPC_ORDER]; int16_t prev_excitation[PITCH_MAX]; - int16_t excitation[PITCH_MAX + FRAME_LEN]; + int16_t excitation[PITCH_MAX + FRAME_LEN + 4]; int16_t synth_mem[LPC_ORDER]; int16_t fir_mem[LPC_ORDER]; int iir_mem[LPC_ORDER]; @@ -267,8 +267,10 @@ static int scale_vector(int16_t *vector, int length) bits = normalize_bits(max, 15); scale = shift_table[bits]; - for (i = 0; i < length; i++) + for (i = 0; i < length; i++) { + av_assert2(av_clipl_int32(vector[i] * (int64_t)scale << 1) == vector[i] * (int64_t)scale << 1); vector[i] = (vector[i] * scale) >> 3; + } return bits - 3; } @@ -592,7 +594,10 @@ static int autocorr_max(G723_1_Context *p, int offset, int *ccr_max, int i; pitch_lag = FFMIN(PITCH_MAX - 3, pitch_lag); - limit = FFMIN(FRAME_LEN + PITCH_MAX - offset - length, pitch_lag + 3); + if (dir > 0) + limit = FFMIN(FRAME_LEN + PITCH_MAX - offset - length, pitch_lag + 3); + else + limit = pitch_lag + 3; for (i = pitch_lag - 3; i <= limit; i++) { ccr = ff_dot_product(buf, buf + dir * i, length)<<1; @@ -967,7 +972,6 @@ static int g723_1_decode_frame(AVCodecContext *avctx, void *data, G723_1_Context *p = avctx->priv_data; const uint8_t *buf = avpkt->data; int buf_size = avpkt->size; - int16_t *out; int dec_mode = buf[0] & 3; PPFParam ppf[SUBFRAMES]; @@ -975,6 +979,7 @@ static int g723_1_decode_frame(AVCodecContext *avctx, void *data, int16_t lpc[SUBFRAMES * LPC_ORDER]; int16_t acb_vector[SUBFRAME_LEN]; int16_t *vector_ptr; + int16_t *out; int bad_frame = 0, i, j, ret; if (!buf_size || buf_size < frame_size[dec_mode]) { @@ -995,8 +1000,8 @@ static int g723_1_decode_frame(AVCodecContext *avctx, void *data, av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); return ret; } - out= (int16_t*)p->frame.data[0]; + out = (int16_t *)p->frame.data[0]; if (p->cur_frame_type == ACTIVE_FRAME) { if (!bad_frame) @@ -1079,7 +1084,7 @@ static int g723_1_decode_frame(AVCodecContext *avctx, void *data, memcpy(p->prev_excitation, p->excitation + FRAME_LEN, PITCH_MAX * sizeof(*p->excitation)); } else { - memset(out, 0, sizeof(int16_t)*FRAME_LEN); + memset(out, 0, FRAME_LEN * 2); av_log(avctx, AV_LOG_WARNING, "G.723.1: Comfort noise generation not supported yet\n"); return frame_size[dec_mode]; @@ -1094,13 +1099,18 @@ static int g723_1_decode_frame(AVCodecContext *avctx, void *data, 0, 1, 1 << 12); memcpy(p->synth_mem, out + FRAME_LEN, LPC_ORDER * sizeof(int16_t)); - if (p->postfilter) + if (p->postfilter) { formant_postfilter(p, lpc, out); + } else { // if output is not postfiltered it should be scaled by 2 + for (i = 0; i < FRAME_LEN; i++) + out[LPC_ORDER + i] = av_clip_int16(out[LPC_ORDER + i] << 1); + } memmove(out, out + LPC_ORDER, sizeof(int16_t)*FRAME_LEN); p->frame.nb_samples = FRAME_LEN; - *(AVFrame*)data = p->frame; - *got_frame_ptr = 1; + + *got_frame_ptr = 1; + *(AVFrame *)data = p->frame; return frame_size[dec_mode]; } |